Estimating Undetected COVID-19 Infections
We specify and estimate a time-varying Markov model of COVID-19 cases for the US in 2020. We find that the estimated level of undetected infections spiked in March and remained elevated through May. However, since late April estimated undetected infections have generally declined though it was not until June or July that detected cases exceeded the estimated number of undetected cases. Our results suggest that the substantial increase in testing capacity in the US has identified a higher percentage of infections. However, these findings also indicate that much of the increase in the number of positive tests since spring represents a true increase in new cases as opposed to an increase resulting from more testing. According to our estimation, more than 20% of the US population has been infected by the Covid-19 virus which is consistent with other published estimates. One shortcoming of our analysis is that we are not able to condition our estimates on the age of people infected or hospitalized which may cause us to underestimate the current number of undetected cases.